Abstract
Objective:
To assess the association between commonly obtained endoscopic and serologic data and bronchoalveolar lavage pepsin assay (BAL) results in children with chronic cough.
Study Design:
We performed a retrospective chart review of seventy-two children with a BAL pepsin obtained through our Aerodigestive Center over an 18-month period. BAL outcomes include evidence of viral, bacterial, or fungal infection, presence of lipid laden macrophages, and cytology (eosinophils, neutrophils, and lymphocytes). Gastrointestinal outcomes include esophagogastroduodenoscopy (EGD) and pH impedance probe findings. Other characteristics include serum eosinophil, neutrophils, and lymphocytes; spirometry; FeNO; and IgE.
Results:
Seventy-two patients underwent BAL pepsin testing. Median age was 4.9 years, 30.6% had severe persistent asthma, and 59.2% were on reflux medication. There was an association between positive BAL pepsin assay and positive viral panel (p=0.002) or fungal culture (p=0.027). No significant association found between positive BAL bacterial culture; BAL cytology; presence of BAL lipid laden macrophages; IgE; spirometry; FeNO; CBC neutrophil, eosinophil, or lymphocytes; pH impedance testing parameters; or EGD pathology.
Conclusions:
BAL pepsin is associated with a positive BAL viral PCR or fungal culture. Lack of correlation between pepsin-positivity and pH-impedance parameters or EGD pathology suggests microaspiration may be due to an acute event (such as a respiratory infection) rather than chronic GERD. This may be especially true in the presence of a positive viral panel or fungal culture when a BAL pepsin is obtained.
Keywords: pediatric pulmonology, pediatric gastroenterology, pediatric otolaryngology, reflux-related microaspiration, aerodigestive model
INTRODUCTION
Chronic cough in children is defined by the American College of Chest Physicians as a cough lasting longer than four weeks (1). Pediatric chronic cough represents a large burden for patients and their families, and a major impact on healthcare delivery. Studies have shown up to 80% of these patients had more than five doctor visits in a year (2), and patients with chronic cough consistently report a lower quality of life (3). Additionally, over nine billion dollars are spent annually on over-the-counter cough and cold medications in the US (4). The cause of chronic cough in children is broad, and diagnosis can be especially challenging if the patient is unable to provide a detailed history of their symptoms or perform spirometry.
Gastroesophageal reflux disease (GERD) is one etiology, among many, that is considered in young children with chronic cough that does not respond to standard therapies (5, 6). Although GERD is a common cause of chronic cough in adults (7, 8), it is considered a much less prevalent cause in children(9). At least two mechanisms have been proposed to explain how reflux can trigger cough. One way involves stimulation of the vagus nerve by acid in refluxate, increasing the cough sensitivity of susceptible patients (10). The other mechanism is refluxate being regurgitated into the larynx or pharynx with possible aspiration into the lower respiratory tract, directly precipitating a cough (10, 11). Cough itself can trigger reflux (12), with the idea that the chronic nature of GERD-associated cough may be due to the interplay between the two (13).
Bronchoalveolar lavage (BAL) can help identify the contribution of reflux to chronic cough, specifically through the use of enzyme-linked immunosorbent assay (ELISA) assays to identify the presence of pepsin. Pepsin is an enzyme solely produced by the chief cells of the stomach, thus its appearance in BAL fluid is suspicious for microaspiration of reflux. BAL pepsin assays are sensitive and specific for aspiration due to gastroesophageal reflux (14), and it has been compared favorably to BAL lipid-laden macrophage index (LLMI) in detecting microaspiration (13). It may also be more helpful in identifying microaspiration than pH probe monitoring (15).
Prior evidence suggests that aspiration occurs more frequently in pediatric patients with viral respiratory infections (16, 17), an acute illness often associated with frequent and severe coughing that can cause increased amount of reflux and may impair airway protective mechanisms, therefore increasing the risk of aspiration. To our knowledge, only a few studies have looked at the association between aspiration, reflux and respiratory infection in children, and their respective conclusions show very little agreement between them. One study of children without previous history of respiratory disease found no association between GERD and viral PCR positivity (18). Another one concluded that children with findings suspicious for reflux-related aspiration were indeed more likely to have a positive viral PCR, if not more likely to have a positive BAL bacterial culture (19). A third study of children with chronic cough or wheezing did report that positive BAL bacterial culture was associated with full-column non-acidic reflux, although no significant association was found between BAL bacterial culture positivity and pepsin assay positivity (20).
In this study, our aim is to see if positive BAL pepsin assay may be associated with other endoscopic and serologic outcomes that we commonly obtain when evaluating children with chronic cough of unclear etiology. Our hypothesis is that evidence of respiratory infection may be found more frequently in patients with a positive pepsin assay.
METHODS
This study was approved by the Institutional Review Board at Children’s National Hospital and all families provided consent for participation (IRB 000001322). We reviewed the charts of children undergoing flexible bronchoscopy and bronchoalveolar lavage through our Aerodigestive Center at Children’s National Hospital (CNH) between May 2015 and September 2018. All children with a BAL pepsin assay were included. Data was collected through the electronic medical record (Cerner Millennium, Cerner Crop., Kansas City, MO, USA). BAL samples were collected per our standard protocol. Briefly, a flexible bronchoscope was inserted into the airway through a laryngeal mask airway or an endotracheal tube. After wedging the bronchoscope into a segmental or subsegmental bronchus, a lavage was performed by instilling 5–10 mL aliquots of sterile normal saline and collected for testing. This was generally repeated 2 or more times in different lobes of the lungs. Between 0.5–2 mL of BAL was collected and stored immediately in ice to be delivered to our pathology department. It was frozen and sent on dry ice to the Clinical Laboratory at Nemours Children’s Clinic in Delaware for analysis by ELISA gastric pepsin A assay, which was performed using previously described methods (21). Per their laboratory, a Pepsin A result greater than 3 units was reported as positive (1 unit is equal to 0.1 ng/mL of Pepsin A). Besides gastric pepsin assay, the BAL was also sent to Children’s National Hospital’s Division of Pathology and Laboratory Medicine accredited clinical laboratory for bacterial culture, fungal culture, respiratory pathogen PCR, and BAL neutrophil, eosinophil, and lymphocyte percentage (22). Each cell type listed on differential is expressed as a percent of total cells visualized on microscopy.
Spirometry results were included for those patients who were able to perform them within 3 months of their bronchoscopy. If there were multiple spirometry attempts during this time, the study performed closest to the date of the bronchoscopy was included. The spirometry was performed with the patient in an upright, seated position in the pulmonary function lab at CNH by a licensed respiratory therapist using MGC Diagnostics equipment with BreezeSuite software v8.5.0.65 (Database Version 750) in accordance to American Thoracic Society recommendations.
Blood was obtained on the day of bronchoscopy for complete blood count with differential and quantitative total IgE. If not performed at time of bronchoscopy, then results closest to bronchoscopy date were included. For serum lymphocytes, we only reported values obtained within a week of the bronchoscopy. All patients undergoing pH impendence testing were asked to discontinue proton pump inhibitors or histamine H2 receptor antagonists five days or 48 hours, respectively, prior to the procedure. A Sandhill pH-MII impendence probe was placed in the distal esophagus about 3–5 cm from the lower esophageal sphincter. Monitoring was performed for 24 hours. Reflux index and John DeMeester scores are used to assess the presence of acid reflux. The DeMeester score was calculated to provide a quantitative measurement of GERD secondary to excessive acid regurgitation. Impedance values showed periods of acid and non-acid reflux in upright and recumbent positions. Excessive non-acid reflux could be identified from the impedance measure. The DeMeester score is based on 1) Longest reflux episode, 2) Number of reflux episodes, 3) Number of reflux episodes > or = 5 minutes, 4) Percent Supine time pH <4, 5) Percent Upright time pH <4, 6) Percent Supine time pH <4. A DeMeester score of >14.7 was determined to be concerning for acid reflux (17). A pH impedance study was considered abnormal if there were 73 or more total reflux episodes per 24 hour period (23). Other pH impedance measures assessed include total reflux episodes, non-acidic episodes, percent of total reflux episodes that were non-acidic, cough correlation symptom index, number of episodes greater than 5 minutes, total number of reflux episodes in the supine position, number of episodes of acid reflux in the supine position, and number of episodes of nonacidic reflux in the supine position.
During EGD, 4–6 biopsy specimens were obtained from the duodenum, 2–4 specimens from the stomach, 2 from the distal esophagus, and 2 from the proximal esophagus. We recorded whether EGD biopsy samples were normal or abnormal on pathology. Results were considered significant for p value < 0.05. Statistical analysis, including post hoc correction for multiplicity (bootstrapping, Monte Carlo) was performed on all data collected using SPSS v25.0 (SPSS, Chicago, IL, USA).
RESULTS
Seventy-two bronchoscopies with BAL pepsin testing were performed during the period of time surveyed. The median age was 4.9 years, 33.3% were female, the median BMI percentile was 62, 59.7% were African American, 30.6% had severe persistent asthma, 59.2% were on reflux medication, 52.7 % had received at least one course of systemic steroid in the year prior to bronchoscopy, and 29.1% were hospitalized in the year prior to bronchoscopy. Fifty-five patients were found to be pepsin-negative on BAL and seventeen patients found to be pepsin-positive. There were thirty-three subjects with completed spirometry that met our study criteria. There were no significant differences between pepsin-negative and pepsin-positive groups with respect to percent predicted FEV1 (median 92.6% vs 88.1%, p-value 0.47), FEV1/FVC absolute ratio (median 0.84 vs 0.86, p-value 0.87) or FEF 25–75% percent predicted (median 82.7% vs 85.2%, p-value 0.98). The median FeNO was 29 parts per billion (ppb) in the negative pepsin group and 7.5 ppb in the positive pepsin group. There were no included comorbidities that were associated with a positive pepsin assay. There were no significant association with any demographic factors and pepsin assay positivity (Table 1).
Table 1.
Demographics
| Characteristic | Overall (n=72) | Negative Pepsin Assay (n=55) | Positive Pepsin Assay (n=17) | P value |
|---|---|---|---|---|
| Age, Median (IQR) | 4.9 (2.2, 9.1) | 4.9 (2.3, 9.9) | 4.2 (1.7, 8.8) | 0.6 |
| Gender, % Female (n) | 33.3% (23) | 32.7% (18) | 35.3% (6) | 0.84 |
| GA, median wk (IQR) | 37 (34, 40) | 37 (34, 40) | 37.5 (30.25, 40) | 0.32 |
| Median BMI, percentile, IQR (n=33) | 62 (35.9, 87.4) | 58.8 (29.7, 84.2) | 69 (46.6, 97.8) | 0.25 |
| Ethnicitya | 0.63 | |||
| African American | 59.7% (43) | 58.2% (32) | 64.7% (11) | |
| Caucasian | 20.8% (15) | 21.8% (12) | 17.6% (3) | |
| Hispanic/Latino | 8.3% (6) | 9.1% (5) | 5.9% (1) | |
| Other | 12.5% (9) | 12.7% (7) | 11.8% (2) | |
| Severe Asthma | 30.6% (22) | 30.9% (17) | 29.4% (5) | 0.75 |
| Urgent visits for Asthma (past 12 months) | 57.4% (31/54) | 54.8% (23/42) | 66.7% (8/12) | 0.46 |
| Systemic steroid courses (past 12 months) | 52.7% (29/55) | 50% (21/42) | 61.5% (8/13) | 0.47 |
| Hospitalized (past 12 months) | 29.1% (16/55) | 28.6% (12/42) | 30.8% (4/13) | 0.88 |
| FeNO, Median (IQR) | 24 (11, 56) | 29 (13, 60) | 7.5 (6.75, 8.25) | 0.26 |
| On oral reflux medication (n=71) | 59.2% (42/71) | 57.4% (31/54) | 64.7% (11/17) | 0.59 |
| Comorbid conditions | ||||
| Rhinorrhea | 52.8% (38/72) | 54.5% (30/55) | 47.1% (8/17) | 0.59 |
| Tonsillar hypertrophy | 26.8% (19/71) | 22.2% (12/54) | 41.1% (7/17) | 0.23 |
| Adenoidal hypertrophy | 66.7% (38/57) | 66.7% (28/42) | 66.7% (10/15) | 0.57 |
| Tonsil and adenoid Hypertrophy | 28.1% (16/57) | 23.8% (10/42) | 40% (6/15) | 0.27 |
| GERD symptoms | 68.6% (48/70) | 64.2% (34/53) | 82.4% (14/17) | 0.16 |
| Neurological Disorderb | 9.72% (7/72) | 9.1% (5/55) | 11.8% (2/17) | - |
| Upper/central airway abnormalityb | 22.2% (16/72) | 18.1% (10/55) | 35.3% (6/17) | 0.14 |
| Lower airway abnormalityb | 6.9% (5/72) | 7.3% (4/55) | 5.9% (1/17) | - |
| Cardiac abnormalityb | 9.7% (7/72) | 10.9% (6/55) | 5.9% (1/17) | - |
| Spirometry | ||||
| Median FEV1, % predicted (IQR) | 92.6 (81, 104.6) | 92.6 (83.4, 105.2) | 88.1 (75.1, 100.1) | 0.47 |
| Median FEV1/FVC (IQR) | 85 (79, 91.5) | 84.5 (78.8, 91.8) | 86 (76.5, 90) | 0.87 |
| Median FEF 25–75, % predicted (IQR) | 83.4 (62.9, 107.2) | 82.7 (67.6, 107.9) | 85.2 (52.0, 113.9) | 0.98 |
One patient was two ethnicities
Further specified in Appendix A
P Value based on Logistic Regression
P Value for continuous variables calculated based on t-test
P Value for binomial variable calculated using Pearson’s χ2 test
BAL outcomes in relation to pepsin assay results are presented in Table 2. A BAL viral panel was positive in 17 of 53 patients (32.1%) in the negative pepsin assay group and 12 of 16 patients (75%) in the positive pepsin group. A BAL fungal culture was positive in 5 of 54 patients (9.3%) in the negative pepsin group and 5 of 16 patients (31.3%) in the positive pepsin group. There was a significant association between positive pepsin assay and positive viral panel (p=0.002) or positive fungal culture (p = 0.027) on BAL; post hoc analysis correcting for multiplicity demonstrated the findings for positive viral panel (p=0.028), but not fungal culture (p=0.33) remained significant. BAL bacterial culture was positive in 10 of 55 patients (18.2%) in the negative pepsin group and 1 of 16 patients (6.3%) in the positive pepsin group. There was no significant association identified between BAL pepsin positivity and a positive BAL bacterial culture (p=0.25). The most common organisms that were identified in the BAL are outlined in Table 2, with further specification in Appendix A (online). The median BAL eosinophil percentage was 0% for both the negative and positive pepsin groups (p=0.14). The median BAL neutrophil percentage was 34% in the negative pepsin group and 62% in the positive pepsin group (p=0.25). The median lymphocyte percentage 13% in the negative pepsin group and 11% in the positive pepsin group (p=0.81). Lipid laden macrophages were present in 81.8% of BAL samples in the negative pepsin assay group and 80% of BAL samples of patients in the positive pepsin assay group (p=0.87).
Table 2.
Bronchoalveolar Lavage (BAL) Studies
| Outcome | Overall (n=72) | Negative Pepsin Assay (n=55) | Positive Pepsin Assay (n=17) | P Value |
|---|---|---|---|---|
| +Viral PCR, % (n=69) | 42% (29/69) | 32.1% (17/53) | 75% (12/16) | 0.002 |
| Rhinovirus/Enterovirus | 27.5% (19/69) | 18.9% (10/53) | 56.3% (9/16) | 0.013 |
| Human Adenovirus | 10.1% (7/69) | 11.3% (6/53) | 6.3% (1/16) | - |
| Coronavirus | 2.9% (2/69) | 0% | 12.5% (2/16) | - |
| Other virusesa | 5.8% (4/69) | 5.7% (3/53) | 6.3% (1/16) | - |
| +Fungal Culture | 14.3% (10/70) | 9.3% (5/54) | 31.3% (5/16) | 0.027 |
| Candida spp. | 11.4% (8/70) | 7.4% (4/54) | 25% (4/16) | - |
| Other fungia | 4.3% (3/70) | 3.7% (2/54) | 6.3% (1/16) | - |
| + Bacterial Culture (n=71) | 15.5% (11/71) | 18.2% (10/55) | 6.3% (1/16) | 0.25 |
| Streptococcus pneumoniae | 9.9% (7/71) | 10.9% (6/55) | 6.3% (1/16) | - |
| Moraxella catarrhalis | 5.6% (4 of 71) | 7.3% (4/55) | 0% | - |
| Haemophilus influenzae | 2.8% (2/71) | 3.6% (2/55) | 0% | - |
| Other bacteriaa | 4.2% (3/71) | 5.5% (3/55) | 0% | - |
| Absolute Eosinophil, Median (IQR) | 0 (0, 2) | 0 (0, 3) | 0 (0, 0) | 0.14 |
| Absolute Neutrophil, Median (IQR) | 42 (10.3, 74.5) | 34 (9, 73) | 62 (25, 79) | 0.25 |
| Absolute Lymphocyte, Median (IQR) | 12.5 (6.0, 22.5) | 13.0 (6.0, 27.0) | 11.0 (5.0, 16.0) | 0.81 |
| +Lipid Laden Macrophages, % (n) | 81.4 (57) | 81.8 (45) | 80 (12) | 0.87 |
Further specified in Appendix A
P Value based on Logistic Regression
P Value for continuous variables calculated based on t-test
P Value for binomial variable calculated using Pearson’s χ2 test
Sixty-two serum cell count differential and fifty-one quantitative IgEs were included in our study (Table 3). There was no significant difference between the pepsin-negative and the pepsin-positive patients in regards to serum eosinophil percentage (median 2.7% vs 2.0%, p= 0.16), neutrophil percentage (median 39% vs 45%, p= 0.7), lymphocyte percentage (median 44.7% vs. 44.6%, p=0.67, or IgE levels (median 41.5 IU/mL vs 31 IU/mL, p= 0.54).
Table 3.
Serum studies
| Outcome | Overall (n=62) | Negative Pepsin Assay (n=48) | Positive Pepsin Assay (n=14) | P value |
|---|---|---|---|---|
| CBC Eosinophil %, Median (IQR) | 2.4 (1.4, 5.5) | 2.7 (1.5, 5.7) | 2.0 (0.7, 3.2) | 0.16 |
| CBC Neutrophil %, Median (IQR) | 40.6 (33.5, 50.3) | 39 (32.9, 50.0) | 45 (33.9, 54.3) | 0.7 |
| CBC Lymphocyte %, Median (IQR | 44.7 (32.8, 53.5) | 44.7 (32.0, 53.5) | 44.6 (33.7, 52.3) | 0.67 |
| Total IgE (IU/mL), Median (IQR) (n=51) | 34 (17, 213) | 41.5 (17.8, 261.8) | 31 (6, 111) | 0.54 |
P Value for continuous variables calculated based on t-test
Twenty-seven patients had a pH-impedance probe study (twenty-two had a negative pepsin assay and five had a positive pepsin assay). There was no significant difference between the pepsin-negative and pepsin-positive groups in any of the assessed pH impedance outcomes (Table 4), including DeMeester Composite score (median 2.6 vs 2.9, p= 0.46), reflux index (median 0.45 vs 0.5, p= 0.44), total number of reflux episodes (49 vs 51, p= 0.96), proportion of patients with an abnormal pH-impedance result as defined by 73 or more episodes of reflux per 24 hours (18.2% vs 20%, p=0.93), number of non-acidic episodes (median 17 vs 16, p= 0.96), percent of total reflux that was non-acidic (median 57% vs. 51%, p= 0.68), cough correlation symptom index (5.4% vs. 17%, p= 0.51), total number of reflux episodes in supine position (median 2 vs 8, p= 0.12), number of episodes of acidic reflux supine (median 1 vs 3.5, p= 0.97), and number of non-acidic reflux supine (1 vs 3.5, p= 0.3). Median number of reflux episodes greater than 5> min was zero for both groups and there was no significant difference between them (p-value 0.42). There were sixty-three EGDs performed. There was no significant difference in the prevalence of abnormal EGD pathology findings between the two groups (52.1% vs 46.7%, p=.71).
Table 4.
GI Endoscopic Outcomes
| Parameter | Overall | Negative Pepsin Assay | Positive Pepsin Assay | P value |
|---|---|---|---|---|
| DeMeester Composite score, IQR (n=27) | 2.9 (1.5, 7.1) | 2.6 (1.3, 8.0) | 2.9 (2.4, 3) | 0.46 |
| Reflux Index, IQR (n = 27) | 0.5 (0.2,1.3) | 0.45 (0.1,0.3) | 0.5 (0.5,0.7) | 0.46 |
| Reflux Index > 4.2 (n=27) | 11.1% (3/27) | 9.1% (2/22) | 20% (1/5) | 0.75 |
| Total Reflux Episodes, IQR (n=27) | 37 (19.5,58) | 49 (14.3, 62.3) | 51 (33, 64) | 0.96 |
| Reflux episodes ≥73, % (n=27) | 18.5 (5/27) | 18.2 (4/22) | 20% (1/5) | 0.93 |
| Non-Acidic Episodes, IQR (n=26) | 16.5 (8.3, 23) | 17 (8,23) | 16 (9, 23) | 0.96 |
| % Non-acidic of Total Reflux, IQR (n=26) | 54 (29.3, 80.3) | 57 (28, 81) | 51 (33, 64) | 0.68 |
| Cough correlation SI, IQR (n=23) | 7% (0, 31.5) | 5.4% (0, 34) | 17% (0, 27) | 0.51 |
| Cough correlation >50% (n=23) | 21.7% (5/23) | 27.8% (5/18) | 0% | 0.18 |
| Episodes greater than >5 min, IQR (n=22) | 0 (0,0) | 0 (0,0) | 0 (0,0) | 0.42 |
| Reflux episodes supine, Acid, IQR (n=23) | 1 (0, 3.5) | 1 (0, 3) | 3.5 (2.8, 5.3) | 0.97 |
| Reflux episodes supine, Nonacid, IQR (n=23) | 2 (0, 4.5) | 1 (0, 4) | 3.5 (2.8, 5.8) | 0.3 |
| Total Reflux Supine, IQR (n=25) | 4 (1, 12) | 2 (0.75, 11.25) | 8 (6, 20) | 0.12 |
| Positive EGD Pathology Findings, % (n) | 50.8% (32/63) | 52.1% (25/48) | 46.7% (7/15) | 0.71 |
P Value for continuous variables calculated based on t-test
P Value for binomial variable calculated using Pearson’s χ2 test
DISCUSSION
BAL pepsin assay has been suggested as a tool to evaluate for reflux-related microaspiration in children with chronic cough. GERD with microaspiration is one possible cause for pepsin deposition in the lung, but there may be other factors that have not been previously identified that may affect pepsin positivity. The aim of our study was to see if the presence of BAL pepsin may be associated with other endoscopic or serologic outcomes we commonly investigate in this population, including evidence of respiratory infection.
No demographic characteristic of our patient population was significantly associated with pepsin positivity. There was no indication that pepsin positive groups had more severe respiratory symptoms, as measured by urgent cares visits, systemic steroid courses, or previous diagnosis of severe asthma. Likewise, there was no significant difference in pulmonary function testing (FEV1, FEV1/FVC, and FEF 25–75%) between the two groups. This is consistent with previous studies assessing the role of BAL pepsin in adult asthmatics that similarly did not find any association between BAL pepsin levels and FEV1 (24).
As previously mentioned, the pediatric literature looking at the association between aspiration, reflux, and respiratory infection is limited to a few studies. Chang, et al reviewed 150 children without previous history of respiratory disease and did not find an association between GERD diagnosed by reflux esophagitis present on distal esophageal biopsy and viral PCR positivity (18). Dr. Calabrese looked at 32 children with chronic respiratory disorders with BAL LLMI and pH assay results and found that children with high LLMI and abnormal esophageal pH testing (designated as “aspirators”) were more likely to have a positive viral PCR (AV, CMV, EBV, EV/RV, PIV, RSV, and Flu A,B) than non-aspirators, but no difference in positive BAL bacterial cultures (19). Rosen et al. performed a prospective, cross-sectional study of 46 children with chronic cough or wheezing, and they found that positive BAL bacterial cultures were associated with full-column, non-acid reflux (20). Dr. Rosen’s study did look at BAL pepsin, but did not find any association between pepsin positivity and a positive BAL bacterial culture. Our results support their findings in respect to the lack of association between bacterial culture and pepsin assay results. While, similar to our study, Calabrese found an association between possible aspiration with laboratory evidence of viral, but not bacterial infection, a key difference is that Calabrese’s “aspirators” had abnormal esophageal pH monitoring while the pH-impedance outcomes in our study did not differ significantly between pepsin-positive and -negative groups.
Viruses and fungi have not been shown to produce pepsin, so the positive pepsin assay results likely represent true microaspiration. The mechanisms behind the increased risk of microaspiration are not entirely clear but may be related to an increase in cough-induced refluxate and/or impairment in swallow. This may be supported by previous studies showing that acute respiratory syncytial virus (RSV) infection in previously healthy infants is associated with increased aspiration and swallowing impairment (16, 17). The swallowing impairment resolved once the infants had fully recovered (17). It’s unclear why in our study a viral PCR, but not a bacterial culture, would be associated with a positive BAL pepsin, considering that both would presumably represent an infection associated with coughing. One possible reason may be that patients with a true bacterial respiratory infection were self-excluded from our study. Considering that bronchoscopies in our study were performed as outpatient procedures, patients would be expected to be close to their baseline state of health on the day of the procedure. Therefore, bronchoscopy would likely have been canceled and rescheduled for patients with fever, dyspnea, exam findings concerning for pneumonia, or cough significantly worse than baseline, as what would be expected with a true bacterial respiratory infection. Thus, these patients would have essentially excluded themselves from our study. On the other hand, patients with recent viral respiratory infection may have fully recovered or may not have even have symptoms significantly different from baseline, so bronchoscopy would have proceeded as scheduled.
As far as we are aware, this is the first study to identify an association between positive BAL pepsin and positive BAL fungal culture, although we recognize that post-hoc analysis suggests these findings may be due to chance. Literature describing fungal organisms and aspiration is scarce. Aspiration of Candida-laden oropharyngeal secretions leading to pneumonia has been rarely reported and mainly limited to elderly intensive care unit patients or patients with cancer (25, 26). Of the ten patients in our study that had a positive fungal culture, eight of them grew Candida albicans. The other two consisted of aspergillus (which is known to colonize the respiratory tract of patients with asthma and cystic fibrosis) (27) and paecilomyces, (which is generally thought to be a contaminant of clinical specimens but can cause pneumonia in immunocompromised individuals) (26). The possible mechanism explaining why a candidal respiratory infection/colonization could be associated with reflux-related microaspiration is unclear. An increased cough related to fungal respiratory infection may induce more reflux, thus providing greater opportunity for aspiration, although, again, candida respiratory infections are rare and generally not seen in relatively healthy children. A more fitting explanation, which could also explain our post-hoc analysis findings, may be that the candida on fungal culture reflects oropharynx colonization rather than a true lower respiratory tract infection. Candida is an organism found normally in oropharynx, and its overgrowth can be a side effect of inhaled corticosteroid deposition on mucous membranes. As most of the patients in our study were on inhaled corticosteroids, they certainly are at higher risk of having greater candida colonization of the oropharynx, which may subsequently be deposited into the lungs by reflux that has traveled up to the larynx or pharynx prior to it being aspirated. This may be supported by the observation that all 4 patients in the pepsin positive group found to have Candida on fungal culture also had a positive viral PCR, as a virus would be a more common cause of cough in children than a fungal respiratory infection (28). Ultimately, further studies are warranted to elucidate the relationship between the positive fungal respiratory cultures and BAL pepsin assays.
In our study, we did not find any significant association between pepsin positivity and pH impedance parameters or EGD pathology findings. Discordance between traditional measures of reflux and BAL pepsin has been previously demonstrated. In a study involving 52 wheezy infants undergoing pH impedance monitoring, EGD and esophageal biopsy, Abdallah et al found elevated pepsin levels in wheezy patients with normal impedance studies compared to healthy controls (15). In another study of children with chronic cough or asthma who underwent pH-impedance testing, EGD, and bronchoscopy, Rosen also found that BAL pepsin did not correlate with any aspect of pH impedance except for number of non-acid events (29). We found that non-acid reflux burden was not associated with pepsin positivity, which is different from Rosen’s study. The cough correlation index was also low and not significantly different in both groups. Aspiration that occurred before, but not during, the time that the MII-pH monitoring was taking place may explain these findings. Also, pathology findings characteristic of GERD may be absent with non-acidic reflux or reflux that does not occur persistently over a prolonged period of time. Perhaps a more fitting explanation for the discordance between pepsin and BAL findings is a temporary increase in reflux and aspiration triggered by an acute event, such as a respiratory infection, occurring prior to flexible bronchoscopy and endoscopic evaluation/pH impedance probe placement.
Airway neutrophilic, lymphocytic, or eosinophilic inflammation, reflux, and aspiration in children have been explored previously. Sacco, et al. found that children with respiratory symptoms and abnormal esophageal pH monitoring had a higher BAL neutrophil percentage and LLM index, but similar eosinophil percentage compared to healthy controls (30). Chang’s study involving children without respiratory symptoms found BAL neutrophil percentage was actually higher in the GERD-negative group (18). In another study evaluating children with chronic lung disease, Starosta et al. found that BAL neutrophil percentage did not correlate with any pH impedance measurement (31). In one study that did directly address the association between pepsin and BAL cellularity in adults with chronic cough due to suspected reflux, there was a trend toward increased neutrophil percentage in the pepsin-positive group, although it did not reach the threshold to be statistically significant (32). There was no significant association between eosinophil or lymphocyte percentage and the various measurements of reflux or aspiration in any of the aforementioned studies. In our study, we did see a trend toward increased BAL neutrophil percentage in the pepsin positive group that did not reach statistical significance, with similar eosinophil and lymphocyte percentages in both groups. Similarly, we did not find any significant differences in serum neutrophil, eosinophil, or lymphocyte percentages between the pepsin-negative and pepsin-positive group. The trend toward higher BAL neutrophil percentage may be consistent with the higher frequency of viral PCR positivity seen in the positive pepsin group as increased neutrophil percentage on BAL has been associated with acute viral respiratory infections (33). Inflammation secondary to aspirated reflux may also explain the trend towards higher BAL neutrophil percentage in the pepsin-positive group. This is supported by previous literature demonstrating an association in patients with cystic fibrosis between increased BAL pepsin and higher levels of BAL IL-8, which activates and attracts neutrophils to areas of inflammation (34).
The percentage of BAL samples that were positive for lipid laden macrophages on pathology were not significantly different between the positive and negative pepsin group, which is not entirely unexpected when reviewing the literature. While an elevated lipid laden macrophage index has been suggested as an indicator of reflux-induced aspiration, further studies have questioned its sensitivity and specificity (35). It also has not been found to correlate well with BAL pepsin in previous studies of reflux and aspiration (14, 15). However, in our study, we were only able to identify whether lipid macrophages were present or not in the BAL sample, rather than calculating a true LLMI, which may make direct comparisons to this literature difficult.
Our study has several limitations. For one, while our population demographics were equally distributed between the pepsin-negative and pepsin-positive groups, the pepsin-positive group had a smaller number of patients. The presence or history of certain medical conditions that may increase the risk or aspiration, thus increasing the likelihood of having a positive pepsin assay, were identified in both groups and outlined in Appendix B (online). There were no obvious differences between the two groups. There were only a small number of pH impedance studies, which would make it difficult to discern statistically significant difference between the two groups with respect to impedance outcomes. Furthermore, as a retrospective study, we are not able to definitively determine the relationship between BAL pepsin positivity and acute viral/fungal infections. We should also mention that the alveolar Type 2 cell in the lung expresses pepsinogen C, which is another protease (36). Although unlikely, we cannot say with definite certainty that this would not affect the results of the BAL pepsin assay, but we would have expected to find this in all samples.
CONCLUSION
Our findings show that BAL pepsin is associated with a positive BAL viral PCR and fungal culture. Considering the lack of correlation between pepsin-positivity and any pH-impedance parameter or abnormal EGD pathology, the presence of pepsin in the airway may represent microaspiration due to an acute event (such as a respiratory infection) rather than chronic GERD. Still further studies are needed to better elucidate any mechanism between BAL pepsin and evidence of respiratory infection, as well as identify when BAL pepsin levels may normalize after acute viral infections and/or aspiration in children with, and without, EGD and pH-impedance confirmed reflux.
Supplementary Material
Acknowledgments
Supported by the National Institutes of Health (NIH 1U01EB021986 [to D.K.P.])
This work was presented in abstract form at the American Thoracic Society’s 2019 International Conference in Dallas, Texas (doi: 10.1164/ajrccm-conference.2019.199.1_meetingabstracts.a1182). It was also presented at the 2019 Annual Aerodigestive Conference in Denver, Colorado, where it won first place in the research abstract competition.
List of abbreviations:
- GERD
gastroesophageal reflux disease
- BAL
bronchoalveolar lavage
- GI
gastrointestinal
- EGD
esophagogastroduodenoscopy
- LLMI
lipid-laden macrophage index
Footnotes
We have no conflicts of interest to disclose. I, Dr. Corey Martin, wrote the first draft of the manuscript. No honorarium, grant or other form of payment was given to anyone to produce the manuscript.
Data for our manuscript is available as individual deidentified participant data.
REFERENCES
- 1.Chang AB, Glomb WB. Guidelines for evaluating chronic cough in pediatrics: ACCP evidence-based clinical practice guidelines. Chest. 2006;129(1 Suppl):260S–83S. Epub 2006/01/24. doi: 10.1378/chest.129.1_suppl.260S. [DOI] [PubMed] [Google Scholar]
- 2.Marchant JM, Newcombe PA, Juniper EF, Sheffield JK, Stathis SL, Chang AB. What is the burden of chronic cough for families? Chest. 2008;134(2):303–9. Epub 2008/07/22. doi: 10.1378/chest.07-2236. [DOI] [PubMed] [Google Scholar]
- 3.Chang AB, Robertson CF, Van Asperen PP, Glasgow NJ, Mellis CM, Masters IB, et al. A multicenter study on chronic cough in children : burden and etiologies based on a standardized management pathway. Chest. 2012;142(4):943–50. Epub 2012/03/31. doi: 10.1378/chest.11-2725. [DOI] [PubMed] [Google Scholar]
- 4.OTC Sales by Category 2016-2019. Consumer Healthcare Products Association. [Google Scholar]
- 5.Orenstein SR, Orenstein DM. Gastroesophageal reflux and respiratory disease in children. J Pediatr. 1988;112(6):847–58. Epub 1988/06/01. doi: 10.1016/s0022-3476(88)80204-x. [DOI] [PubMed] [Google Scholar]
- 6.Naime S, Batra SK, Fiorillo C, Collins ME, Gatti M, Krakovsky GM, et al. Aerodigestive Approach to Chronic Cough in Children. Curr Treat Options Pediatr. 2018;4(4):467–79. doi: 10.1007/s40746-018-0145-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chung KF, Pavord ID. Prevalence, pathogenesis, and causes of chronic cough. Lancet. 2008;371(9621):1364–74. Epub 2008/04/22. doi: 10.1016/S0140-6736(08)60595-4. [DOI] [PubMed] [Google Scholar]
- 8.Irwin RS. Chronic cough due to gastroesophageal reflux disease: ACCP evidence-based clinical practice guidelines. Chest. 2006;129(1 Suppl):80S–94S. Epub 2006/01/24. doi: 10.1378/chest.129.1_suppl.80S. [DOI] [PubMed] [Google Scholar]
- 9.Chang AB, Oppenheimer JJ, Weinberger M, Grant CC, Rubin BK, Irwin RS, et al. Etiologies of Chronic Cough in Pediatric Cohorts: CHEST Guideline and Expert Panel Report. Chest. 2017;152(3):607–17. Epub 2017/06/25. doi: 10.1016/j.chest.2017.06.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kahrilas PJ, Smith JA, Dicpinigaitis PV. A causal relationship between cough and gastroesophageal reflux disease (GERD) has been established: a pro/con debate. Lung. 2014;192(1):39–46. Epub 2013/11/14. doi: 10.1007/s00408-013-9528-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ravelli AM, Panarotto MB, Verdoni L, Consolati V, Bolognini S. Pulmonary aspiration shown by scintigraphy in gastroesophageal reflux-related respiratory disease. Chest. 2006;130(5):1520–6. Epub 2006/11/14. doi: 10.1378/chest.130.5.1520. [DOI] [PubMed] [Google Scholar]
- 12.Paterson WG, Murat BW. Combined ambulatory esophageal manometry and dual-probe pH-metry in evaluation of patients with chronic unexplained cough. Dig Dis Sci. 1994;39(5):1117–25. Epub 1994/05/01. doi: 10.1007/BF02087567. [DOI] [PubMed] [Google Scholar]
- 13.Ing AJ, Ngu MC, Breslin AB. Pathogenesis of chronic persistent cough associated with gastroesophageal reflux. Am J Respir Crit Care Med. 1994;149(1):160–7. Epub 1994/01/01. doi: 10.1164/ajrccm.149.1.8111576. [DOI] [PubMed] [Google Scholar]
- 14.Farrell S, McMaster C, Gibson D, Shields MD, McCallion WA. Pepsin in bronchoalveolar lavage fluid: a specific and sensitive method of diagnosing gastro-oesophageal reflux-related pulmonary aspiration. J Pediatr Surg. 2006;41(2):289–93. Epub 2006/02/17. doi: 10.1016/j.jpedsurg.2005.11.002. [DOI] [PubMed] [Google Scholar]
- 15.Abdallah AF, El-Desoky T, Fathi K, Elkashef WF, Zaki A. Clinical Utility of Bronchoalveolar Lavage Pepsin in Diagnosis of Gastroesophageal Reflux among Wheezy Infants. Can Respir J. 2016;2016:9480843. Epub 2016/08/16. doi: 10.1155/2016/9480843. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hernandez E, Khoshoo V, Thoppil D, Edell D, Ross G. Aspiration: a factor in rapidly deteriorating bronchiolitis in previously healthy infants? Pediatr Pulmonol. 2002;33(1):30–1. Epub 2001/12/18. doi: 10.1002/ppul.10022. [DOI] [PubMed] [Google Scholar]
- 17.Khoshoo V, Edell D. Previously healthy infants may have increased risk of aspiration during respiratory syncytial viral bronchiolitis. Pediatrics. 1999;104(6):1389–90. Epub 1999/12/10. doi: 10.1542/peds.104.6.1389. [DOI] [PubMed] [Google Scholar]
- 18.Chang AB, Cox NC, Purcell J, Marchant JM, Lewindon PJ, Cleghorn GJ, et al. Airway cellularity, lipid laden macrophages and microbiology of gastric juice and airways in children with reflux oesophagitis. Respir Res. 2005;6:72. Epub 2005/07/19. doi: 10.1186/1465-9921-6-72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Calabrese F, Rizzo S, Giacometti C, Panizzolo C, Turato G, Snijders D, et al. High viral frequency in children with gastroesophageal reflux-related chronic respiratory disorders. Pediatr Pulmonol. 2008;43(7):690–6. Epub 2008/05/27. doi: 10.1002/ppul.20841. [DOI] [PubMed] [Google Scholar]
- 20.Rosen R, Johnston N, Hart K, Khatwa U, Katz E, Nurko S. Higher rate of bronchoalveolar lavage culture positivity in children with nonacid reflux and respiratory disorders. J Pediatr. 2011;159(3):504–6. Epub 2011/07/23. doi: 10.1016/j.jpeds.2011.05.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.He Z, O’Reilly RC, Bolling L, Soundar S, Shah M, Cook S, et al. Detection of gastric pepsin in middle ear fluid of children with otitis media. Otolaryngol Head Neck Surg. 2007;137(1):59–64. Epub 2007/06/30. doi: 10.1016/j.otohns.2007.02.002. [DOI] [PubMed] [Google Scholar]
- 22.Patregnani JT, Brooks BA, Chorvinsky E, Pillai DK. High BAL sRAGE is Associated with Low Serum Eosinophils and IgE in Children with Asthma. Children (Basel). 2020;7(9). doi: 10.3390/children7090110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Shay S, Tutuian R, Sifrim D, Vela M, Wise J, Balaji N, et al. Twenty-four hour ambulatory simultaneous impedance and pH monitoring: a multicenter report of normal values from 60 healthy volunteers. Am J Gastroenterol. 2004;99(6):1037–43. Epub 2004/06/08. doi: 10.1111/j.1572-0241.2004.04172.x. [DOI] [PubMed] [Google Scholar]
- 24.Hunt EB, Ward C, Power S, Sullivan A, Pearson JP, Lapthorne S, et al. The Potential Role of Aspiration in the Asthmatic Airway. Chest. 2017;151(6):1272–8. Epub 2017/03/17. doi: 10.1016/j.chest.2017.03.005. [DOI] [PubMed] [Google Scholar]
- 25.Haron E, Vartivarian S, Anaissie E, Dekmezian R, Bodey GP. Primary Candida pneumonia. Experience at a large cancer center and review of the literature. Medicine (Baltimore). 1993;72(3):137–42. Epub 1993/05/01. [PubMed] [Google Scholar]
- 26.Schnabel RM, Linssen CF, Guion N, van Mook WN, Bergmans DC. Candida pneumonia in intensive care unit? Open Forum Infect Dis. 2014;1(1):ofu026. Epub 2015/03/04. doi: 10.1093/ofid/ofu026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Agarwal R. Allergic bronchopulmonary aspergillosis. Chest. 2009;135(3):805–26. Epub 2009/03/07. doi: 10.1378/chest.08-2586. [DOI] [PubMed] [Google Scholar]
- 28.Everard ML. Paediatric respiratory infections. Eur Respir Rev. 2016;25(139):36–40. Epub 2016/03/02. doi: 10.1183/16000617.0084-2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Rosen R, Johnston N, Hart K, Khatwa U, Nurko S. The presence of pepsin in the lung and its relationship to pathologic gastro-esophageal reflux. Neurogastroenterol Motil. 2012;24(2):129–33, e84–5. doi: 10.1111/j.1365-2982.2011.01826.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sacco O, Silvestri M, Sabatini F, Sale R, Moscato G, Pignatti P, et al. IL-8 and airway neutrophilia in children with gastroesophageal reflux and asthma-like symptoms. Respir Med. 2006;100(2):307–15. Epub 2005/07/09. doi: 10.1016/j.rmed.2005.05.011. [DOI] [PubMed] [Google Scholar]
- 31.Starosta V, Kitz R, Hartl D, Marcos V, Reinhardt D, Griese M. Bronchoalveolar pepsin, bile acids, oxidation, and inflammation in children with gastroesophageal reflux disease. Chest. 2007;132(5):1557–64. Epub 2007/10/11. doi: 10.1378/chest.07-0316. [DOI] [PubMed] [Google Scholar]
- 32.Grabowski M, Kasran A, Seys S, Pauwels A, Medrala W, Dupont L, et al. Pepsin and bile acids in induced sputum of chronic cough patients. Respir Med. 2011;105(8):1257–61. Epub 2011/05/20. doi: 10.1016/j.rmed.2011.04.015. [DOI] [PubMed] [Google Scholar]
- 33.McNamara PS, Ritson P, Selby A, Hart CA, Smyth RL. Bronchoalveolar lavage cellularity in infants with severe respiratory syncytial virus bronchiolitis. Arch Dis Child. 2003;88(10):922–6. Epub 2003/09/23. doi: 10.1136/adc.88.10.922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.McNally P, Ervine E, Shields MD, Dimitrov BD, El Nazir B, Taggart CC, et al. High concentrations of pepsin in bronchoalveolar lavage fluid from children with cystic fibrosis are associated with high interleukin-8 concentrations. Thorax. 2011;66(2):140–3. doi: 10.1136/thx.2009.130914. [DOI] [PubMed] [Google Scholar]
- 35.Krishnan U, Mitchell JD, Tobias V, Day AS, Bohane TD. Fat laden macrophages in tracheal aspirates as a marker of reflux aspiration: a negative report. J Pediatr Gastroenterol Nutr. 2002;35(3):309–13. doi: 10.1097/00005176-200209000-00013. [DOI] [PubMed] [Google Scholar]
- 36.Foster C, Aktar A, Kopf D, Zhang P, Guttentag S. Pepsinogen C: a type 2 cell-specific protease. Am J Physiol Lung Cell Mol Physiol. 2004;286(2):L382–7. Epub 2003/10/28. doi: 10.1152/ajplung.00310.2003. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.